Abstract Boehmite and diaspore are the two economic ore minerals of karstic bauxites. Although their genesis has been studied from different perspectives, their formation mechanism is controversial. Random forest (RF) model of machine learning was employed to extract the combining characteristics of trace elements in boehmite-type bauxite (BTB) and diaspore-type bauxite (DTB). The BTB predominantly exhibits higher median concentrations of Co, Ni, V, and Cr, while the DTB exhibits a more significant enrichment in elements of U, Hf, Th, and Zr. Combining the characteristics of La/Yb and Ga/Al ratios, it is found that disparities between BTB and DTB are mirroring to those between basic rocks and intermediate-felsic rocks. Furthermore, the Zr-Cr-Ga diagram and Ni/Zr versus Cr/Zr plot were utilized to examine the correlation between karstic bauxite (BTB or DTB), lateritic bauxite, and their respective weathered parent rocks. It is found that BTB exhibits consistent characteristics with lateritic bauxite weathered from basic rock and its parent rocks. Similarly, DTB displays consistent characteristics with lateritic bauxite weathered from intermediate-felsic rock and its parent rocks. Through studying the relationship between Ni content and Fe3+/Fe2+ ratios, it has been discovered that the presence of trace elements like Ni in source materials can affect or regulate ore-forming process, ultimately driving the transformation of gibbsite into either boehmite or diaspore. Consequently, the formation of BTB and DTB is significantly influenced, or even governed, by the composition of their source materials. Our study initially highlights a novel insight into the significant impact of source material’s geochemical composition on the formation of boehmite and diaspore, which is associated with weathered igneous rocks categories.
Read full abstract